The Known Good Updated 25 Jul 2026
DeepSeek: DeepSeek V3 vs Microsoft: Phi 4 ·DeepSeek vs Microsoft

DeepSeek: DeepSeek V3 vs Microsoft: Phi 4 — head-to-head comparison

DeepSeek: DeepSeek V3 profile Microsoft: Phi 4 profile All comparisons
DeepSeek: DeepSeek V3 — DeepSeek
Quality
#44 of 61
44based on 3 of 6 evaluations
Known Good Index
Speed
#280 of 316
20.9
Output tokens per second
Price
#105 of 340
$0.35
Blended per 1M tokens
In $0.20 · Out $0.80
Cache hit price
#10 of 206
$0.003
per 1M tokens (−99%)
Context
#225 of 345
164k
tokens (~300 pages)
Microsoft: Phi 4 — Microsoft
Quality
#50 of 61
33based on 3 of 6 evaluations
Known Good Index
Speed
#142 of 316
56.1
Output tokens per second
Price
#40 of 340
$0.09
Blended per 1M tokens
In $0.07 · Out $0.14
Cache hit price
unranked
—
per 1M tokens
Context
#334 of 345
16k
tokens (~25 pages)
Summary

On the Known Good Index, DeepSeek: DeepSeek V3 scores 44 and Microsoft: Phi 4 scores 33 — a 11-point lead for DeepSeek: DeepSeek V3. They share 10 published evaluations, with DeepSeek: DeepSeek V3 ahead on 9 and Microsoft: Phi 4 ahead on 1.

Blended pricing is $0.35 per 1M tokens for DeepSeek: DeepSeek V3 and $0.09 for Microsoft: Phi 4 — Microsoft: Phi 4 is about 4.0× cheaper.

Throughput is 21 tokens per second for DeepSeek: DeepSeek V3 and 56 for Microsoft: Phi 4, with time to first token at 1.1s and 0.2s. DeepSeek: DeepSeek V3 accepts 164k tokens of context against 16k.

DeepSeek: DeepSeek V3 is open weights under Open weights (restricted use); Microsoft: Phi 4 is open weights under Open weights (unrestricted). 3 tracked hosts serve one or both, and 1 serve both.

Generated from tracked data on 25 Jul 2026 · how these comparisons are computed →

At a glance
MetricDeepSeek: DeepSeek V3Microsoft: Phi 4
Known Good Index 44 based on 3 of 6 evaluations 33 based on 3 of 6 evaluations
Blended $/1M $0.35 $0.09
Output tok/s 21 56
Context window 164k 16k
Providers 3 1
Jump to section

Quality

Known Good Index and the capability indices, side by side

IndicesComponents
Index scores

Each index is min-max normalised across tracked models and scaled 0–100 from published evaluations. We do not run these evaluations.

MetricDeepSeek: DeepSeek V3Microsoft: Phi 4
Known Good Index 44 33
Mathematics Index 40 39
Science Index 53 53
Reasoning Index 52 47
Openness Index 60 60
The Known Good

Quality Breakdown

Every evaluation both models report a published score for

Evaluations both models report

Bar = DeepSeek: DeepSeek V3 · marker = Microsoft: Phi 4 · scores ingested from Epoch AI (CC-BY 4.0), LiveBench (Apache-2.0)

GPQA Diamond57 / 56
MATH / AIME16 / 14
LiveBench60 / 42
LiveBench Coding62 / 31
LiveBench Data Analysis61 / 45
LiveBench Instruction Following75 / 58
LiveBench Language47 / 26
LiveBench Mathematics61 / 42
LiveBench Reasoning57 / 48
MATH Level 565 / 65

Right-hand figures read DeepSeek: DeepSeek V3 / Microsoft: Phi 4. DeepSeek: DeepSeek V3 leads on 9 of 10, Microsoft: Phi 4 on 1.

The Known Good

Arena Elo

Human preference rating from blind pairwise votes

Arena ratings

Elo from blind pairwise votes, ingested from Arena. Higher is better.

MetricDeepSeek: DeepSeek V3Microsoft: Phi 4
Text Arena 1,333 1,217
Text Arena (style controlled) 1,358 1,256
The Known Good

Price & Cost

Published API pricing by token type

Per token typeCache discount
Published pricing

USD per 1M tokens. Blended is a 3:1 input:output mix. Live from OpenRouter, cross-checked against models.dev. Lower is better.

MetricDeepSeek: DeepSeek V3Microsoft: Phi 4
Input $/1M $0.20 $0.07
Output $/1M $0.80 $0.14
Blended $/1M $0.35 $0.09
Cache read $/1M $0.003 —
The Known Good
We publish blended price rather than cost-per-task. Cost-per-task requires token counts from an evaluation suite we do not run. Blended price is computed from published rates and is fully reproducible — see the pricing methodology.

Context Window

Maximum input accepted and maximum output emitted

Context and output limits

Provider-declared limits, from metadata via models.dev and OpenRouter. Higher is better.

MetricDeepSeek: DeepSeek V3Microsoft: Phi 4
Context window 164k 16k
Max output tokens 16k 16k
The Known Good

Speed & Latency

Median throughput and time to first token across hosting providers

Output speedTime to first token
Throughput and latency

Medians across all tracked providers, derived from OpenRouter endpoint telemetry.

MetricDeepSeek: DeepSeek V3Microsoft: Phi 4
Output tokens/s 21 56
Time to first token 1.1s 0.2s
The Known Good

Providers

Which hosts serve each model, and at what price and speed

3 hosts · 1 serve both
Full provider benchmarks →
Provider DeepSeek: DeepSeek V3 blendedDeepSeek: DeepSeek V3 tok/s Microsoft: Phi 4 blendedMicrosoft: Phi 4 tok/s
StreamLake $0.35 24 — —
DeepInfra $0.46 20 $0.09 70
Novita $0.62 14 — —

Per-host pricing and telemetry from OpenRouter. A dash means that host does not serve that model.

The Known Good
Sources for this page
Pricing & providers — OpenRouter Metadata — models.dev Benchmarks — Epoch AI (CC-BY 4.0)Benchmarks — LiveBench (Apache-2.0)Elo — Arena Full attribution →